The high consumption of wine in France is thought to be an important dietary factor in the low incidence of coronary heart disease (CHD) mortality and has been suggested at least in part to provide a possible explanation for the phenomenon known as the "French Paradox" (Renaud & De Lorgeril 1992), France being an exception compared with most other countries because CHD mortality is low despite a high intake of saturated fat.
There is a considerable literature on the alleged beneficial effects of red wine in relation to prevention of coronary heart disease (CHD). Epidemiological data suggest the protection afforded by wine is superior to that of other alcoholic beverages such as beer and spirits, indicating that factors other than alcohol content in wine is contributing to the effect (St Leger et al., 1979; Renaud & De Lorgeril 1992). In a prospective study in Copenhagen, Denmark various parameters (including alcohol intake, smoking habit and body mass index) were assessed in 13,285 people succeeded by a 12 year follow-up of mortality. It was shown that low to moderate intake of wine (but not beer or spirits) was associated with lower mortality from cardiovascular and cerebrovascular diseases and other causes (Gronbaek et al., 1995). These results confirmed those previously reported in the USA (Klatsky & Armstrong, 1993).
There is growing evidence that the free radical chain reaction of lipid peroxidation involving the oxidation of low density lipoproteins (LDL) plays an important contributory role in the development of atherosclerosis and CHD (Steinberg, 1993).
Frankel et al (Lancet 1993 341, 454-457) examined the ability of dilute, dealcoholised red wine to inhibit the oxidation of human LDL in vitro, and found the wine to be very active as an antioxidant. The authors suggested that the routine consumption of red wine may "reduce oxidation of lipoproteins and reduce thrombotic phenomena". However, the authors admitted that "we need to know more about the pharmaco-kinetics of wine flavonoids and the absorption and metabolism of wine phenols . . . if we are to evaluate further the potential role of antioxidant compounds in red wine in reducing CHD".
Flavonoids belong to a group of substances called polyphenols (PP), so named because they contain two or more phenolic groups. Polyphenols occur abundantly in red wine and consist of a large number of different chemical substances of varying molecular weights. The chief polyphenol components of grapes and wine, and their concentrations, are described by Shahidi & Nazck (1995) in "Food phenolics: sources, chemistry, effects and applications" (Technomic Publishing Co., Lancaster Pa., USA) p136-146. Among the polyphenols are the following classes: flavonoids (a term often used to denote polyphenols in general, but more commonly in Europe to denote only the flavones), the flavanols, proanthocyanidins (also called procyanidols, procyanins, procyanidins and tannins) and anthocyanins.
The flavones are compounds with a basic structure shown in FIG. 2 in which two benzene rings (A and B) are linked with a heterocyclic six member ring C containing a carbonyl group. Ring B can be joined in position 2 (as illustrated) to give a flavone or to position 3 to give an iso flavone. Hydroxylation can occur at positions 3, 5, 7 and 3', 4', 5' to give compounds called flavonols. Typical examples of flavonols are: quercetin, (hydroxylated at positions 3, 5, 7, 3', 4'), kaempferol (hydroxylated at positions 3, 5, 7, 4'), and myricetin (hydroxylated at positions 3, 5, 7, 3', 4', 5'). They can exist naturally as the aglycone or as O-glycosides (e.g. D-glucose, galactose, arabinose, rhamnose etc). Other forms of substitution such as methylation, sulphation and malonylation are also found.
The flavanols have a basic structure shown in FIG. 3. The two most common flavanols are catechin (hydroxyl groups positions 5, 7, 3', 4') and its stereo-isomer epi-catechin. The hydroxyl groups can be esterified with gallic acid (shown in FIG. 4). The proanthocyanidins are polymers of catechin and/or epicatechin and can contain up to 8 units or more.
The anthocyanins are coloured substances with a basic structure shown in FIG. 5. They are sometimes called anthocyanidins. Typical examples are: cyanidin (hydroxylated at positions 3, 5, 7, 3', 4'), delphinidin (hydroxylated at positions 3, 5, 7, 3', 4', 5') and pelargonidin (hydroxylated at positions 3, 5, 7, 3'). The hydroxyl groups are usually glycosylated and/or methoxylated (e.g. malvidin at 3', 5').
Within the general term "polyphenols" are included the dihydroxy- or tri-hydroxy benzoic acids and the phytoalexins, a typical example of which is resveratrol (shown in FIG. 6).
The most widely used method for the determination of LDL oxidation is to employ the transition metal copper (specifically Cu.sup.2 + ions) as a catalyst to promote the oxidation of endogenous lipid hydroperoxides. Antioxidants present in LDL, especially alpha tocopherol, delay the oxidation process and produce a so called lag phase. The process can be easily followed in a UV spectrophotometer because the oxidation reaction produces conjugated dienes which can be continuously monitored at 234 nm (Esterbauer et al., 1989). To preserve LDL from oxidation during storage, EDTA is added to complex copper and other trace elements. This excess EDTA interferes with the copper catalysed oxidation. EDTA can be removed by dialysing the LDL preparation before addition of the copper ions or an excess of copper ions can be added to compensate for those complexed with EDTA.
The results of in vitro experiments somewhat similar to those described by Frankel et al., (Lancet 1993, cited above) were also reported by Frankel et al in 1995 (J. Agricult. and Food Chemistry 43, 890-894). The authors of this publication draw attention to the difficulty of interpreting in vitro data. Thus "Although the phenolic compounds have similar chemical properties, their reducing capacity is not a very precise predictor of their antioxidant activity. In the LDL oxidation assay and other tests for antioxidant activity, the system is typically heterogeneous and physical properties, such as lipophilicity, solubility and partition between the aqueous and lipid phases of LDL can become important in determining antioxidant activity".
Indeed, those skilled in the art appreciate that extrapolation from in vitro findings to in vivo situations is frequently inappropriate. As an example, the reader is referred to the publication of McLoone et al, (1995 Proc. Nutr. Soc. 54, Abstract 168A), which shows that although the compound lutein has the potential to inhibit LDL oxidation in vitro, supplementation of the diet of human volunteers with lutein for 2 weeks (which gave a 6-fold increase in the levels of lutein in plasma) had no effect on LDL oxidation.
Some in vivo trials have been conducted to investigate the possible health benefits of red wine. Fuhrman et al, (1995 Am. J. Clin. Nutr. 61, 549-554) found that "some phenolic substances that exist in red wine, but not in white wine, are absorbed, bind to plasma LDL, and may be responsible for the antioxidant properties of red wine" and provided, in their words, the first demonstration "that red wine consumption inhibits the propensity of LDL to undergo lipid peroxidation", and that this may contribute to attenuation of atherosclerosis. However, a study by Sharpe et al, the results of which were published (Q.J. Med. 1995 88, 101-108) nearly contemporaneously with those of Fuhrman et al, found that neither consumption of red wine nor white wine had any effect "on total cholesterol, triglycerides, HDL or measures of antioxidant status, including the susceptibility of LDL to oxidation".
De Rijke et al. also investigated the matter and conducted a randomized double-blind trial. They reported their findings in 1996 (Am. J. Clin. Nutr. 63, 329-334) and stated that "The results of this study do not show a beneficial effect of red wine consumption on LDL oxidation".
Thus, to summarise, there are several reports that dilute red wine can inhibit LDL oxidation in in vitro assays, but that these findings cannot necessarily be extended to the in vivo situation. Further, the in vivo data relating to inhibition of LDL oxidation by red wine consumption are at best conflicting and there is no clear evidence to suggest that red wine consumption has any effect on LDL oxidation.
A number of compositions are now publicly available which are prepared from wine or grape by-products and which may contain polyphenols (albeit at quite low levels in some of the compositions). Among these are French Paradox capsules (available from Arkopharma). French Paradox capsules are made by preparing an extract from marc (the grape skin waste remaining after wine fermentation). Most of the polyphenols present in the grape skins are alcohol-soluble, and so tend to be extracted into the fermenting wine. Hence, French Paradox capsules have actually rather low polyphenol content. (Other publicly available compositions include an anthocyanin-containing powder (obtainable from Sefcal) made from a grape skin extract, and which is used as a food colourant, and a proanthocyanidin-containing composition ("Endotelon") prepared from grape seeds.)
Even if French Paradox capsules contained significant amounts of polyphenols, it is not clear that oral consumption of such a synthetic polyphenol composition would exert the same therapeutic effect allegedly associated with red wine consumption. For example, as explained by Goldberg (1995 Clin. Chem. 41, 14-16) the alcohol content of wine keeps polyphenols in solution in wine and in the human intestine, such that they might be available for absorption. A synthetic, alcohol-free polyphenol powder may be completely ineffective because the polyphenols are insufficiently soluble in the intestine (in the absence of alcohol) to be absorbed. Additionally, absorption into the bloodstream may not be sufficient for any anti-oxidant effect to be exerted on LDL--intimate association of the polyphenols with the LDL fraction may be required.
It is recognised that many diseases are caused or provoked by a free radical oxidation mechanism e.g. cancer, cataracts, diabetes etc. Antioxidant nutrients such as vitamin E, vitamin C and others are thought to prevent free radical oxidation in many organs and tissues. Thus the absorption of polyphenols which are effective antioxidants are likely to have an effect on free radical/oxidation diseases in general, and the use of polyphenols may be much wider than a treatment or prevention of coronary heart disease.
Nevertheless, CHD is one of the major causes of mortality and morbidity in the western world, and therefore of particular interest. Pathogenesis of the condition consists essentially of a two stage process involving first the development of atherosclerotic plaques and then formation of a thrombus (clot) on the plaque (a process called thrombosis) which may cause arterial occlusion, the consequences of which can be myocardial infarction (MI) and sudden death. Other diseases which are caused by thrombosis are stroke and venous thrombosis. The initial stage in the formation of a thrombus is the aggregation of platelets which then release coagulation factors into the blood causing the production of fibrin clots. Once the blood clots are formed they can be removed by a process known as fibrinolysis, which is essentially the dissolution of clots and the degradation of fibrin to degradation products. Thus there are at least two processes by which thrombosis can be prevented: inhibiting the aggregation of platelets, or increasing fibrinolysis.
Abnormal vascular smooth muscle cell (VSMC) proliferation may contribute to the formation of obstructive lesions in coronary heart disease, atherosclerosis, restenosis, stroke and smooth muscle neoplasms of the bowel and uterus, uterine fibroid or fibroma.
It has been known for many years that TGF-.beta. is one of the most potent cell growth inhibitors (Massague, 1990), and several authors have found that TGF-.beta. inhibits VSMC proliferation (Assolian, 1986; Bjorkerud, 1991; Owens, 1988; Kirschenlohr, 1993). Human VSMC produce TGF-.beta. in a latent, inactive form which is activated proteolytically by the serine protein plasmin, which in turn is obtained from plasminogen by a family of plasminogen activators (PAs), such as tissue plasminogen activator (tPA) (Lyons, 1990). An increase in total plasma TGF-.beta. is considered effective in inhibiting the growth of VSMC, since the latent form is converted into the active form by plasmin.
Several authors have developed methods for the estimation of the plasma level of TGF-.beta. and to search for pharmaceutical compounds which may stimulate TGF-.beta. production both in the latent and active form. In U.S. Pat. No. 5,545,569 (Grainger et al) a method is claimed for determining in vitro the effectiveness of compounds which increase the plasma level of TGF-.beta. and stimulate its production using the techniques described therein. W094/26303 (Grainger et al) discloses a method for maintaining or increasing the vessel lumen diameter in a diseased or injured vessel of a mammal by administering an effective amount of TGF-.beta. activator or production stimulator. The compound Tamoxifen (trans-2 [4 (diphenyl-1-butanyl)phenoxy]-dimethyl ethylamine is claimed as being effective, since it stimulates the production of TGF-.beta., and increases the ratio of active to latent TGF-.beta.. Another compound showing activity is aspirin (Grainger, et al, 1995) which increases both total and active serum TGF-.beta. in normal people but only total TGF-.beta. in patients with coronary heart disease.
An increase in platelet aggregation has also been significantly associated with prevalence (Elwood et al, 1991) and incidence (Thaulou et al, 1991) of CHD. Platelet aggregation is conveniently studied using a platelet aggregometer in which a suspension of platelets freshly obtained from blood is placed in contact with an agonist which causes aggregation. Many agonists may be used but the most typical are arachidonic acid, ADP, collagen and thrombin. From a measurement of the maximum aggregation (%) it is possible to study the effects of the inhibitors of platelet aggregation which may be given orally or by injection to the subject. One of the most effective substances in preventing platelet aggregation is aspirin, which inhibits cyclo-oxygenase activity and formation of thromboxane, a necessary factor in thrombus formation (Moncada & Vane, 1979). Aspirin also prevents CHD, stroke and sudden death (Hennekens et al, 1988).
The fibrinolytic system constitutes a cascade of extra-cellular proteolytic reactions tightly regulated by activators and inhibitors. The enzyme tissue-type plasminogen activator (t-PA) converts plasminogen to plasmin, which in turn dissolves fibrin clots. t-PA is a glycoprotein synthesized in the endothelial cells, which is adsorbed on to fibrin in order to be activated. Plasminogen activator inhibitor (PAI)-1, is a serine protease inhibitor and acts as a specific inhibitor of t-PA. PAI-1 exists in three forms: active, latent, and as an inactive complex. It is synthesised in endothelial cells, liver and platelets.
In the circulation most tPA (95%) is complexed with PAI-1. Very little tPA and PAI-1 are in the free (active) form. A decreased fibrinolytic activity is thought to be due to an increase in PAI-1 level or activity, which results in decreased activation of plasminogen to plasmin by tPA. This is important because of reports of an association between decreased fibrinolytic activity and risk of CHD (Mehta et a!, 1987) and MI. Impaired fibrinolysis, mainly due to elevation of plasma PAI-1, is a common finding in thrombotic disease. In the Northwick Park Heart Study, a prospective epidemiological study of middle aged men (40-54 at entry), Meade et al (1987) reported that a decreased fibrinolytic activity is a major independent risk factor for future CHD. Cross-sectional studies of patients with angina pectoris or previous myocardial infarction have consistently shown a decreased fibrinolytic activity in patients compared to control (Hamsten et al., 1985 and 1986; Johnson 1984; Paramo et al., 1985; Aznar et al., 1986; Francis 1988; and Olofson et al., 1989). PAI-1 concentrations have been shown to be higher in MI patients compared to controls (Hamsten et al., 1987).
Because of the role of platelet aggregation and fibrinolysis in the formation of thrombi, a method of decreasing platelet aggregation and/or increasing fibrinolysis could be employed as a method of treatment of thrombotic diseases in general, and CHD in particular.